A Local Information Advantage

Local Investments Give Mutual Fund Managers the Edge

Research by Tobias J. Moskowitz

Finance theory prescribes holding well-diversified portfolios to reduce the overall risk of your investments. Yet, most people tend to hold poorly diversified portfolios, and heavily favor domestic stocks over international stocks. New research linking geography and mutual fund performance suggests that having the right information makes local investments much more attractive to investors.


In their paper, "The Geography of Investment: Informed Trading and Asset Prices," Tobias J. Moskowitz, assistant professor of finance at the University of Chicago Graduate School of Business, and Joshua D. Coval of Harvard Business School examine the mutual fund industry to determine why people would prefer to invest in stocks of companies located near them.

Professional mutual fund managers are expected to maintain a diversified portfolio of funds by trade, and are not expected to succumb to biases for local stocks. Though most research suggests that mutual fund managers typically fail to beat the market with their investments, Moskowitz and Coval present the first evidence that some managers do better than the market and their competitors, but only in their selection of local stocks.

The authors studied the holdings of mutual funds that were actively managed over the past twenty years and found that fund managers earn higher than expected returns in their local stocks. In particular, the most agile funds, those that are small with few holdings, and managers with the strongest investment community ties (e.g., older managers located in remote areas) exhibit the greatest gains from local investment. Large, young funds from competitive and dense markets exhibit no such performance.

This new finding suggests that fund managers earn higher returns as a result of information they have acquired about local companies, giving them greater ability to select local stocks. The average fund manager generates an additional return of 2.67 percent per year from local investments, defined as holdings within approximately sixty miles of the fund headquarters, relative to out of town holdings. In addition, local stocks avoided by fund managers underperform the local stocks they hold by a risk-adjusted 3 percent per year.

Managers with the best local stock selection abilities, and presumably the most information, tilt their portfolios as much as 20 to 25 percent toward local stocks, reaping the largest gains from local investment. Local stocks held by these fund managers outperform out of town holdings by as much as 3 percent per year.

"We don't necessarily believe that mutual fund managers have an exceptional ability to pick stocks. It's rather that some fund managers are gaining access to private information that others can't," says Moskowitz.

Connections that Count

The authors first matched the location of every fund manager to the headquarters of the stocks they owned, and then tracked each fund's quarterly holdings from 1975 to 1994. They measured the degree of local bias by determining the dollars actually invested in local stocks relative to the potential dollar investment that exists locally. The remaining holdings were classified as distant stocks.

While their results suggest a method of identifying informed investors, the authors caution that the success of local managers is difficult for outsiders to replicate.

"If I can see that a fund manager in Arkansas is buying stocks A, B, and C, and selling D, E, and F, I can mimic that," says Moskowitz. "But funds only report their positions every three to six months, as required by the SEC. So by the time I try to duplicate all their local positions, they could be six months old, and I won't make any profits."

A variety of fund characteristics affect stock selection ability, including total asset value, numbers of holdings, and age of the fund. Moskowitz and Coval also classified funds by metropolitan location: large cities (funds in the 20 most populated cities); small cities (funds not located in any of the 20 most populated cities); and remote cities (funds located at least 150 miles from any of the 20 most populated cities).

The funds with the largest assets show less local bias than small funds, and unlike small funds, do not produce higher returns on local investments. Similar results emerge from comparisons of funds according to the number of holdings. These results support the hypothesis that information plays a key role in local stock selection ability, because small funds with few holdings are better able to monitor local information and pursue active trading strategies.

The oldest funds in the study also weigh more heavily in favor of local stocks, and their local holdings outperform out of town holdings. The youngest funds exhibit no such bias, and show less ability to select well-performing local companies. Moreover, funds from large cities show only a modest amount of local bias in their holdings, while funds from small, remote cities show a large local bias and perform above average in their local investments. The more established community ties and relationships of older managers in remote locations with few competitors provide these managers with a local information advantage.

Although part of this local advantage may be access to private information from local firms, proximity to local firms also lowers the travel, time, and research costs associated with obtaining information. Investors located near a firm can visit its operations, talk to suppliers and employees, as well as assess the local market conditions in which the firm operates.

"What's really driving these effects is perhaps a handful of managers in the smaller markets who have close connections with the local business community," says Moskowitz. "We think of it as a 'country club effect,' since in a small community, mutual fund managers may run in the same circles as CEOs of local companies, and some information may move back and forth."

Why Not Go Local?

Overall, the study shows that fund managers with superior ability to identify promising local stocks focus their investments locally, while funds with no such ability choose to hold a more geographically diverse portfolio.

Why don't funds hold even more local stocks? Given that local investments perform better, Moskowitz and Coval suggest that fund managers should specialize entirely in local stocks, allowing investors to diversify across all the different local mutual funds. However, the structure of the mutual fund industry is such that fund managers' careers are dependent on asset growth and large clienteles. This provides little incentive to focus exclusively in small local markets. Moskowitz suggests that the mutual fund manager who starts out in a small city and becomes successful is likely to move to New York to work for a large mutual fund rather than staying and capitalizing on local information and connections.

Regarding the evidence of local bias among mutual fund managers, Moskowitz adds: "People who hold mutual funds might want to find out how well diversified those portfolios really are."

Tobias J. Moskowitz is assistant professor of finance at the University of Chicago Graduate School of Business.

>> Request a copy of the research paper featured in this article